Today we successfully implemented a joystick into our GUI for manually controlling our vehicle. We were also able to stream joystick data to our cRIO which then can be processed into motor commands. Here is a screenshot of our GUI recognizing the Joystick and streaming it’s data.

Joystick and all of its components

In our last update on Driving with Tentacles we got the tentacles calculated but we still needed to calculate the classification area for each tentacle. Classification area is the area that our vehicle will be covering as it drives along the tentacle. In order to choose the correct tentacle to drive by we must know if our vehicle will have a clear path along the tentacle. So here it is, the classification area for three tentacles in the image below. We show only three because otherwise it would be very difficult to see the difference between each tentacle. “S” stands for the classification area and a “.” represents an actual tentacle.

3 Tentacles and their Classification Areas

Throughout the research for path calculation methods the team came across various ways of calculating paths including very untraditional one, using tentacles. This method of path calculation was invented by Felix von Hundelshausen, Michael Himmelsbach, Falk Hecker, Andre Mueller, and Hans-Joachim Wuensche and first used applied at both the C-Elrob (European Land Robot Trial) 2007 and the 2007 DARPA Urban Challenge. This method has a very simple concept, it creates number of tentacles extending out from the center of the robot. Each tentacle has a different length and curvature based on the travel speed of the robot. Travel speeds are grouped intro speed sets, each containing different sets of tentacles. “Our approach does not accumulate data and uses a local ego-centered grid, thereby avoiding the SLAM problem.” – Driving with tentacles: Integral structures for sensing and motion One of the reasons we selected this method for path calculation was because it is innovative, was proven to work and is quite easy to implement. The link above will allow you to download the report that the team used to implement tentacles for Prometheus.The image below shows the tentacles calculated for the robot driving at half of the maximum speed (1.1m/s).

81 tentacles (only 40 is shown)

Red dot indicates the front of the robot facing to the right. Since each tentacle is divided into segments that will contain the probability of an object being at the point of the segment we were able to plot our tentacles using the x, y coordinate of each segment. Black lines over dots represent actual tentacles formed by the segments. There is a total of 81 tentacles represented in the map. This is the first version of the tentacles and there is much work to be done tweaking the parameters and experimenting with the results on the robot. Updates to come.